A model evaluation of estimates of breath-to-breath alveolar gas exchange.

A mathematical model has been implemented for evaluation of methods for estimating breath-to-breath alveolar gas exchange during exercise in humans. This model includes a homogeneous alveolar gas exchange compartment, a dead space compartment, and tissue spaces for CO2 (alveolar and dead space). The dead space compartment includes a mixing portion surrounded by tissue and an unmixed (slug flow) portion which is partitioned between anatomical and apparatus contributions. A random sinusoidal flow pattern generates a breath-to-breath variation in pulmonary stores. The Auchincloss algorithm for estimating alveolar gas exchange (Auchincloss et al., J. Appl. Physiol. 21: 810-818, 1966) was applied to the model, and the results were compared with the simulated gas exchange. This comparison indicates that a compensation for changes in pulmonary stores must include factors for alveolar gas concentration change as well as alveolar volume change and thus implies the use of end-tidal measurements. Although this algorithm yields reasonable estimates of breath-to-breath alveolar gas exchange, it does not yield a "true" indirect measurement because of inherent error in the estimation of a homogeneous alveolar gas concentration at the end of expiration.